Low Acetaldehyde Collection Efficiencies for 24-Hour Sampling with 2

DNPH derivatization–HPLC during the Sydney Particle Study, 2012: a contribution ... Steven Sai Hang Ho , Judith C. Chow , John G. Watson , Ho Sa...
0 downloads 0 Views 131KB Size
Environ. Sci. Technol. 2007, 41, 580-585

Low Acetaldehyde Collection Efficiencies for 24-Hour Sampling with 2,4-Dinitrophenylhydrazine (DNPH)-Coated Solid Sorbents JASON S. HERRINGTON,† ZHI-HUA (TINA) FAN,‡ PAUL J. LIOY,‡ AND J U N F E N G ( J I M ) Z H A N G * ,‡,§ Joint Graduate Program in Exposure Science, University of Medicine and Dentistry of New Jersey (UMDNJ)sRobert Wood Johnson Medical School (RWJMS) and Rutgers, The State University of New Jersey, 683 Hoes Lane West, Piscataway, New Jersey 08854, Environmental and Occupational Health Sciences Institute (EOHSI), Exposure Science Division, UMDNJ-RWJMS and Rutgers, The State University of New Jersey, 170 Frelinghuysen Road, Piscataway, New Jersey 08854, UMDNJ-School of Public Health, 170 Frelinghuysen Road, Piscataway, New Jersey 08854

Airborne aldehyde and ketone (carbonyl) sampling methodologies based on derivatization with 2,4-dinitrophenylhydrazine (DNPH)-coated solid sorbents could unequivocally be considered the “gold” standard. Originally developed in the late 1970s, these methods have been extensively evaluated and developed up to the present day. However, these methods have been inadequately evaluated for the long-term (i.e., 24 h or greater) sampling collection efficiency (CE) of carbonyls other than formaldehyde. The current body of literature fails to demonstrate that DNPH-coated solid sorbent sampling methods have acceptable CEs for the long-term sampling of carbonyls other than formaldehyde. Despite this, such methods are widely used to report the concentrations of multiple carbonyls from long-term sampling, assuming ∼100% CEs. Laboratory experiments were conducted in this study to evaluate the long-term formaldehyde and acetaldehyde sampling CEs for several commonly used DNPHcoated solid sorbents. Results from sampling known concentrations of formaldehyde and acetaldehyde generated in a dynamic atmosphere generation system demonstrate that the 24-hour formaldehyde sampling CEs ranged from 83 to 133%, confirming the findings made in previous studies. However, the 24-hour acetaldehyde sampling CEs ranged from 1 to 62%. Attempts to increase the acetaldehyde CEs by adding acid to the samples post sampling were unsuccessful. These results indicate that assuming ∼100% CEs for 24-hour acetaldehyde sampling, as commonly done with DNPH-coated solid sorbent methods, would substantially under estimate acetaldehyde concentrations. * Corresponding author phone: (732) 445-0158; fax: (732) 4450116; e-mail: [email protected]. † Joint Graduate Program in Exposure Science. ‡ Environmental and Occupational Health Sciences Institute. § UMDNJ-School of Public Health. 580

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 2, 2007

Introduction As part of the United States (U.S.) Environmental Protection Agency’s (EPA) national air toxics monitoring program, hazardous air pollutants (HAPs) are measured at 23 National Air Toxics Trends Stations (NATTS) in 22 cities (1). In addition, state and local agencies are monitoring HAPs at over 300 air toxics monitoring stations nationwide (1). Formaldehyde and acetaldehyde are two of the volatile organic compounds (VOCs) typically monitored at these sites. Formaldehyde and acetaldehyde concentrations are usually derived from 24hour time-integrated samples collected with 2,4-dinitrophenylhydrazine (DNPH)-coated solid sorbent methods based on EPA compendium method TO-11A (2). The goal of the air toxics monitoring program is to provide data that elucidates spatial and temporal trends, supports exposure assessments, and aids in the evaluation of air quality models; all of which is ultimately aimed at supporting the reduction of public exposure to HAPs such as formaldehyde and acetaldehyde (3). Therefore, the 24-hour time-integrated formaldehyde and acetaldehyde concentrations reported from air toxics monitoring sites must be accurate and reliable. Originally developed in the late 1970s, DNPH-coated solid sorbent sampling methods have been extensively evaluated and developed up to present day. The first documented DNPH-coated solid sorbent sampling method appears to be DNPH-coated Amberlite XAD-2 (4). Since then, similar methods have been developed and evaluated by numerous researchers (e.g., refs 5-28). However, a large number of these method evaluations were limited to formaldehyde (4, 5, 7, 10, 11, 13, 19, 23). Although method evaluations were extended to other carbonyls (6, 8, 9, 12, 14-18, 20-22, 2428), they were only evaluated for short-term sampling durations that were on the order of minutes to a few hours (6, 8, 9, 14-16, 18, 21, 22, 24, 26-28), with the longest sampling duration being 12 h (12). Through an extensive literature search, we were only able to find three studies during which carbonyls other than formaldehyde were evaluated on DNPHcoated solid sorbents for long-term sampling (i.e., 24 h or greater) (17, 20, 25). Lazarus (25) reported low CEs; and Grosjean (17), and Grosjean and Grosjean (20) evaluated breakthrough of the collection media, which does not necessarily reflect CE. The breakthrough tests were conducted using cartridge-impinger or cartridge-cartridge sampling trains, and no breakthrough simply means that the downstream sample did not collect a measurable amount of carbonyl(s). The absence of breakthrough cannot automatically be translated in to ∼100% CE on the upstream cartridge; because CE is defined as the ratio of the carbonyl concentration determined from the collection media, to the actual (known) concentration. In the case of DNPH-coated solid sorbents, CE is largely dependent on the reaction efficiency of the reversible acid-catalyzed carbonyl-DNPH derivatization reaction (8, 21, 24, 27, 29, 30). In fact, Grosjean and Grosjean (21) noted that although there was no breakthrough observed by Grosjean and Grosjean (20), the CEs for DNPHcoated solid sorbents sampled in dry air were low. It is not the intent of this manuscript to be a review of every DNPH-coated solid sorbent sampling method. However, our exhaustive literature search failed to produce any documentation from the body of readily available literature, which appropriately evaluated the long-term sampling CEs of DNPH-coated solid sorbent sampling methods for carbonyls other than formaldehyde. Due to the lack of appropriate evaluations, can we assume that DNPH-coated solid sorbent sampling methods have ∼100% CEs for long-term sampling of carbonyls other than formaldehyde? The goal of 10.1021/es061247k CCC: $37.00

 2007 American Chemical Society Published on Web 12/05/2006

TABLE 1. HPLC-UV Analytical Conditions time (min) 0 5 33 50 55

% A: water/acetonitrile/ tetrahydrofuran 60/30/10 v/v

% B: water/ acetonitrile 60/40 v/v

100 100 0 0 100

0 0 100 100 0

flow rate injection volume detector wavelength

1 mL/min 20 µL 360 nm

the current study was to provide an answer to the above by determining the long-term formaldehyde and acetaldehyde sampling CEs for several commonly found DNPH-coated solid sorbents. A dynamic atmosphere generation system was employed to generate formaldehyde and acetaldehyde gas standards to determine CEs. This is significant, because in 1998, Kleindienst et al. (23) noted that with the exception of a few studies (19 and now 23), relatively few studies had systematically evaluated the performance of DNPH-coated solid sorbent methods with the use of formaldehyde gas standards, and none evaluated long-term sampling of acetaldehyde gas standards. We focused our efforts on formaldehyde and acetaldehyde, because they are ubiquitous in indoor and outdoor air, and are of significant health concerns (In June 2004, formaldehyde was classified as a human carcinogen based on sufficient evidence from epidemiological studies (31, 32). Acetaldehyde has been classified as probable human carcinogen by the U.S. EPA (33). In addition, both aldehydes are potent eye and respiratory tract irritants).

Materials and Methods DNPH-Coated Solid Sorbents. The following commercially available DNPH-coated cartridges were used in this study: SUPELCO’s (Bellefonte, PA) LpDNPH air monitoring cartridge (referenced as SUPELCO in tables and figures); Waters (Milford, MA) Sep-Pak DNPH-silica cartridge (referenced as WATERS in tables and figures); and Waters Sep-Pak XPoSure aldehyde sampler (referenced as XPOSURE in tables and figures). These cartridges were selected based on their ubiquitous citation in the literature. In addition to the commercially available DNPH-coated cartridges, we evaluated an “in house” cartridge (referenced as HOUSE in tables and figures). The DNPH-coated cartridges prepared in house have been reported in detail earlier by Zhang et al. (34) and Zhang and Smith (35). Briefly, C18 Sep-Pak cartridges (Waters Corporation) were freshly coated with twice-re-crystallized DNPH, as per a method adopted from EPA compendium method TO-11A (2). Sample Extraction and Analysis. All samples were extracted and analyzed using the following procedures. All sample cartridges and extracts were stored in the dark at -20 °C. Sampled cartridges were gravimetrically eluted with 4 mL of acetonitrile (ACN). Sample extracts were analyzed using an HPLC system (Spectra Physics P4000 mobile phase pump, Spectra Physics AS3000 autosampler, Spectra Physics UV2000 programmable UV detector; and Waters Nova-Pak C18 column (3.9 × 150 mm, 60 Å, 4 µm) and guard cartridge (Nova-Pak, 4 µm, 60 Å, C18 Guard-Pak)). The use of the analytical program, as described in Table 1, was able to clearly resolve the formaldehyde- and acetaldehyde-DNPH derivatives from all other carbonyl-DNPH derivatives. Carbonyl concentrations were determined through calibration curves prepared using commercially available standard

FIGURE 1. Dynamic atmosphere generation system for generation of test atmospheres with carbonyl compounds solutions of pure carbonyl-DNPH derivatives purchased from SUPELCO.

Experiments Sample Breakthrough. Prior to conducting any CE experiments for short-term and long-term sampling CE experiments, we estimated an appropriate sampling flow rate based on molar stoichiometry and the carbonyl concentrations in our dynamic atmosphere generation system. This would allow for sufficient collection (approximately 10 times greater than the blank concentrations) of the carbonyls present in our dynamic atmosphere generation system (Figure 1) (36, 37), while avoiding sample breakthrough. Based on the concentrations in our dynamic atmosphere generation system, and our longest sampling period of 48 h at a nominal flow rate of 150 mL min-1; the consumption of DNPH, based on molar stoichiometry, would be ∼16 and ∼32% for the commercial and in house cartridges, respectively. This is well below the manufacturers’ loading recommendations based on 50% consumption. Therefore, there should be a sufficient amount of DNPH to promote the forward carbonyl-DNPH derivatization reaction to completion. In addition, we conducted sample breakthrough experiments to demonstrate any correlations, or lack thereof, with CE. Similar to Grosjean and Grosjean, (20) we sampled the four types of DNPH cartridges with cartridge-cartridge sampling trains for 24 h at a nominal flow rate of 150 mL min-1. For all of the experiments the sampling flow rate was regulated by an SKC adjustable low flow regulator (Houston, TX) and the sampling flow rate was verified at the start and end of each experiment with a DryCal DC-Lite primary flow controller (Bios International Corp., NJ). The mean sampling flow rate accuracy was ( 6.5%. Collection Efficiency. A dynamic atmosphere generation system (Figure 1) used in our previous laboratory studies generated atmospheres of formaldehyde and acetaldehyde gas standards (36, 37). Formaldehyde and acetaldehyde were generated with permeation devices (VICI Metronics, WA). Once equilibrated at their operating temperature, the permeation devices were found to be constant over a period of several months. The concentrations delivered were determined by the mass delivered from the permeation device and the total flow rate of the dynamic atmosphere generation system (Figure 1). Concentrations were cross checked by collecting samples with the Passive Aldehydes and Ketones (PAKS) method (36, 37). Using this system, we determined the formaldehyde and acetaldehyde sampling CEs for DNPHcoated solid sorbents, for both short-term and long-term sampling (Table 3). Short-Term CE. We wanted to reproduce the CE results, reported in studies referenced earlier, for short-term sampling VOL. 41, NO. 2, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

581

TABLE 2. Dynamic Atmosphere Generation System Parameters parameter

mean ( SD

oven no. 1 temperature (°C) oven no. 2 temperature (°C) system flow rate (L min-1) formaldehyde concentration (µg m-3) acetaldehyde concentration (µg m-3)

30.4 ( 0.5 29.9 ( 1.2 0.98 ( 0.02 22.8 ( 0.5 47.7 ( 1.1

durations using DNPH-coated solid sorbents (Table 3). This was necessary to confirm previous findings, and serve as a cross-check that our dynamic atmosphere generation system was generating the expected carbonyl concentrations. To reduce costs and analytical time, tests were conducted to determine the short-term CEs of Waters Sep-Pak DNPHsilica cartridges (WATERS) only. Samples were collected for 3 h, at a RH of 30%, and at a nominal flow rate of 150 mL min-1. Long-Term CE. We determined the long-term sampling CEs of all four DNPH-coated solid sorbents. The long-term CEs were determined for 24- and 48-hour sampling durations at 30% RH (Table 3). For these experiments samples were collected at a nominal flow rate of 150 mL min-1. In addition, because Grosjean (1996) observed that DNPH-coated solid sorbents performed poorly at low RH (21), we evaluated the effect of RH on the long-term CEs for a 24-hour sampling duration at 60% RH (Table 3). For these experiments, samples were collected at a nominal flow rate of 75 mL min-1. CE Deviation with Sampling Time/Volume. We attempted to determine at what time, if any, the CE would begin to deviate from ∼100%. For this experiment we collected samples at 2, 6, 12, 16, 18, 20, 22, and 24 h; at a RH of 30%, and at a nominal flow rate of 100 mL min-1. To reduce costs and analytical time, the test was only evaluated for Waters Sep-Pak DNPH-silica cartridges (WATERS). In addition, the sample extracts from this experiment were split; one of the sample extracts was analyzed as it was, while ∼5 µL (one drop) of hydrochloric (HCl) acid was added to the other extract, shaken, allowed to sit for 3 h, and then analyzed. The samples were allowed to sit for 3 h, because the current body of literature suggests that carbonyl-DNPH derivatization reactions are complete after a few minutes to hours, for lower molecular weight carbonyls. This experiment was conducted to evaluate if incomplete derivatization was responsible for a reduction, if any, in CE as the sampling duration increased; and if so, whether the addition of HCl acid would push the carbonyl-DNPH derivatization reaction to completion after sample collection. Stability. Sample instability on the DNPH-coated solid sorbent and/or in the ACN extract has been attributed to the poor performance of DNPH-coated solid sorbents for the collection of other carbonyls (e.g., acrolein) (12). Therefore, we conducted experiments to determine the stability of acetaldehyde on the DNPH-coated solid sorbents and in the ACN extracts; and attempted to assess what role, if any, sample stability had on the determination of CEs. For the cartridge stability experiment we spiked DNPH-coated cartridges with an acetaldehyde standard. These cartridges were then stored at 4 °C. Two of these spiked cartridges were extracted and analyzed at a time on days 1, 2, 3, 4, 7, 11, and 14 post spiking (Figure 2). For the extract stability experiment, several cartridge extracts containing acetaldehyde-DNPH derivatives from the cartridge stability experiment were stored at room temperature (∼25 °C), and reanalyzed at 1, 2, 3, 6, 7, 10, 13, and 15 days post extraction (Figure 3). We did not evaluate the stability of the XPOSURE cartridge due to its similarity with the WATERS cartridge. Inter-Laboratory Quality Check. To assess our analytical accuracy, an independent, outside laboratory provided us (blinded) with extracts, which covered a range of samples, 582

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 2, 2007

field positive controls, standards, and pure ACN. A total of 10 samples were analyzed by our laboratory and the outside laboratory. Our analytical results were then compared with the outside laboratory’s analytical results (Figure 4).

Results and Discussion Breakthrough and Collection Efficiency. Results from our 3-hour CE experiment (Table 3) confirmed that DNPH-coated solid sorbents have ∼100% CEs for formaldehyde and acetaldehyde under short-term sampling durations. This observation is consistent with the findings of other researchers who appropriately defined and determined CE for shortterm sampling with DNPH-coated solid sorbents (5, 10, 19, 23, 27). In addition, results from our 3-hour CE experiments confirmed that our dynamic atmosphere generation system was generating the expected formaldehyde and acetaldehyde concentrations. Results from our 24- and 48-hour (both 30% RH) CE experiments (Table 3) confirmed that DNPH-coated solid sorbents have ∼100% CEs for formaldehyde under long-term sampling durations. For these experiments, the formaldehyde sampling CEs ranged from 83 to 111%. This observation is consistent with the findings of Sirju and Shepson (19) who appropriately defined and determined the formaldehyde (only) CE for 24-hour sampling with DNPH-coated solid sorbents. The results from our 24- and 48-hour (both 30% RH) CE experiments (Table 3); however, indicated that DNPHcoated solid sorbent sampling methods consistently poorly measured acetaldehyde concentrations. For these experiments, the acetaldehyde sampling CEs ranged from 0 to 62%. The CEs among the commercial DNPH cartridges appear to agree well with one another, perhaps only reflecting some minor variability across cartridge type and/or the variability of the dynamic atmosphere generation system. However, the “in house” cartridge (HOUSE) did not perform as well as the commercial cartridges for acetaldehyde and for the 48-hour formaldehyde at 30% RH. The discrepancy between the commercial and “in house” cartridges could possibly be the result of different substrates and/or acids. The commercial cartridges’ substrate is silica gel and the in house cartridges’ substrate is C18. Lazarus (25) and Kleindienst et al. (23) also observed that under certain conditions C18 cartridges would under measure carbonyls relative to silica gel cartridges. Results from our breakthrough and CE experiments indicate that despite low acetaldehyde sampling CEs, acetaldehyde was never found to be in detectable quantities on the second (downstream) cartridges. This observation is consistent with Grosjean and Grosjean (21) observation that the use of two cartridges in series with no breakthrough does not explain low CE. Our results are also consistent with the finding of Kleindienst et al. (23) that a discrepancy between DNPH-coated silica gel and C18 cartridges was not associated with breakthrough. As stated earlier, the absence of breakthrough does not necessarily mean ∼100% CE, because DNPH-coated solid sorbents operate on both adsorption and absorption, and the CE of DNPH-coated solid sorbents is largely dependent on the reaction efficiency of the reversible carbonyl-DNPH derivatization reaction (8, 21, 24, 27, 29, 30). The reversible carbonyl-DNPH derivatization reaction is complex; and is dependent upon parameters such as substrate moisture, substrate pH level, and substrate pH strength. An imbalance in these parameters could result in the incomplete carbonyl-DNPH derivatization reaction. For example, protonation of the carbonyl group at a low pH level will promote the nucleohphilic addition, but concurrently reduces the amount of available un-protonated DNPH (30). Because of these competing effects, the carbonyl-DNPH derivatization reaction rate passes through a maximum at a characteristic pH level (30). Based on our results, the formaldehyde-DNPH derivatization reaction is not adversely

TABLE 3. Collection Efficiency, Ratio of Concentration Measured to Concentration Generated in the Dynamic Atmosphere Generation System, Reported as Mean ( Sd, Parentheses Represent Sample Number experimental condition 3 h at 30%

RHa,b

24 h at 30%

RHa,b

48 h at 30% RHa,b 24 h at 60% RHa,d a

Temperature ) 30 °C.

b

carbonyl

SUPELCO

WATERS

XPOSURE

HOUSE

formaldehyde acetaldehyde formaldehyde acetaldehyde formaldehyde acetaldehyde formaldehyde acetaldehyde

89 ( (3) 93 ( 8c (3) 83 ( 4 (3) 39 ( 7 (3) 89 ( 8 (3) 51 ( 22 (3) 101 ( 8 (3) 27 ( 4 (3)

87 ( 11 (3) 43 ( 3 (3) 93 ( 4 (3) 43 ( 2 (3) 101 ( 13 (3) 29 ( 2 (3)

111 ( 4 (3) 62 ( 7 (3) 105 ( 19 (3) 40 ( 11 (3) 121 ( 32 (3) 30 ( 2 (3)

104 ( 25 (3) 1 ( 2 (3) 14 ( 8 (3) 0 (3) 133 ( 27 (3) 9 ( 2 (3)

Sample flow rate ) 150 mL min-1.

10c

c

Only determined with WATERS cartridge.

d

Sample flow rate ) 75 mL min-1.

FIGURE 4. Extract acetaldehyde stability. N ) 2 for each data point and Y error bars represent the difference among duplicate samples.

FIGURE 2. Acetaldehyde CE deviation for untreated extracts and extracts treated with HCl acid. N ) 2 for each data point and Y error bars represent the difference among duplicate samples.

FIGURE 3. Cartridge acetaldehyde stability. N ) 2 for each data point and Y error bars represent the difference among duplicate samples. affected by longer sampling durations; however, the acetaldehyde-DNPH derivatization reaction is adversely affected, perhaps due to an imbalance as previously discussed. We speculate that the larger sample volumes associated with the longer sampling durations upsets the substrate pH level and/or strength, possibly due to evaporation of the acid catalyst. However, it is possible that the larger sample volumes caused the DNPH to react with the acid to form a salt, as proposed by Grosjean and Grosjean (21). The mechanisms behind our observations need further evaluation. Results from the CE deviation experiment (Figure 2) indicate that the acetaldehyde CE drops off gradually from

2 to 16 h of sampling, and then the CE appears to stabilize from 16 to 24 h. Figure 2 represents the trend in acetaldehyde CE for our sampling conditions (i.e., temperature, relative humidity, etc.). A decrease or increase in sampling rate, temperature, relative humidity, and/or carbonyl concentration should alter the CE trend. It is worth noting that the apparent tapering off in CE is suggestive that the CE reaches some sort of equilibrium. This observation is consistent with the observation between 24- and 48-hour sampling, where there was no decrease from 24- to 48- hour sampling. Again, illustrating the complex relationship between long-term sampling and CE. The split extracts from this experiment that were treated with HCl acid were not able to provide an increase in CE, and in some cases, the CE appeared to decrease slightly. In addition to the split extracts, several samples were treated with HCL acid on the cartridge post sampling (results not shown). This was done in an attempt to take advantage of increased reaction efficiency afforded by the cartridge surface area; however, the results from this experiment were also unsuccessful in increasing the acetaldehyde CE. Perhaps it would have been more appropriate to add a base, because as Grosjean and Grosjean proposed (21), the sampling medium may have become too acidic. We can only speculate, as these mechanisms are expected to be complex and warrant future investigations. Future investigations would need to evaluate substrate pH level and strength (with various acids and/or pH buffers) on the cartridge preand post-sampling, and other reaction parameters (e.g., time, temperature, etc.). Grosjean and Grosjean (21) observed that DNPH-coated solid sorbents performed poorly when sampled under relatively low RH (3-7%). The authors reasoned that lower RH levels resulted in the DNPH-coated solid sorbent becoming too acidic when sampled in dry air, and possibly resulted in the formation of a salt from a reaction with DNPH and the acid; thereby reducing the reaction efficiency of the carbonyl-DNPH derivatization reaction. Although the RH in our first two long-term CE experiments (30%) was not VOL. 41, NO. 2, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

583

FIGURE 5. Interlaboratory quality extremely low or unreasonable, we conducted 24-hour CE experiments at 60% RH (Table 3) to examine the effect of RH on the long-term sampling CEs for DNPH-coated solid sorbents. Results from the 24-hour CE experiments at 60% RH (Table 3) continue to indicate that DNPH-coated solid sorbent sampling methods have ∼100% CEs for formaldehyde, but substantially under measure acetaldehyde. For these experiments, the formaldehyde sampling CEs ranged from 101 to 133% and the acetaldehyde sampling CEs ranged from 1 to 30%. We observed that the acetaldehyde sampling CEs decreased with an increase in humidity, which is contrary to what Grosjean (17), and Grosjean and Grosjean (21) observed. However, it is difficult to make any direct comparisons with Grosjean (17), and Grosjean and Grosjean (21) observations, because both of these studies evaluated the performance of DNPH-coated C18 cartridges; and our DNPHcoated C18 cartridges virtually had not detected acetaldehyde for any of our CE experiments. Regardless, our results indicate that an increase in RH decreases the efficiency of DNPHcoated silica gel cartridges. The mechanism for this remains unclear and needs to be evaluated in future studies. In an attempt to examine sample instability as a possible cause for our observed low acetaldehyde CEs for long-term sampling with DNPH-coated solid sorbents, we examined the stability of the acetaldehyde-DNPH derivatives on the collection media and in the ACN extracts. Results from the stability experiments (Figures 3 and 4) indicate that our observed low acetaldehyde CEs were not associated with sample instability on the DNPH-coated solid sorbents or in the ACN extracts. Results from our cartridge experiment (Figure 3) indicate that acetaldehyde was stable on the cartridge within (20% over 14 days. It should be noted that the cartridge stability results reflect analytical and cartridge spiking variability. Results from our extract experiment (Figure 4) indicate that the acetaldehyde-DNPH derivative was stable in the extract within (20% over 15 days post extraction. It is important to note that although our sample stability experiments were based on spiked samples, we did not observed any changes in the acetaldehyde-DNPH derivative both on cartridge and in extract over time for “real” samples. Furthermore, these results are consistent with the findings of other researchers (10, 12, 16, 18). These results suggest that our low acetaldehyde CEs were not associated with sample instability. Finally, results from our inter-laboratory quality (Figure 5) indicate that our analytical results are consistent with the results obtained by an outside laboratory. Results from linear least-squares regression on the two sets of analytical results demonstrates a slope value close to 1, an intercept value close to 0, and a high R2. This indicates that our analysis of carbonyl-DNPH derivatives was robust and was not a source of error. 584

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 2, 2007

Other Evidence of Low Acetaldehyde CE. Previous work conducted in our laboratory also provides evidence that DNPH-coated solid sorbent sampling methods have low acetaldehyde CEs for long-term sampling. In the Relationships of Indoor, Outdoor, and Personal Air (RIOPA) study (38), the 48-hour formaldehyde concentrations determined by DNPH-coated solid sorbents appear to agree reasonably well with the formaldehyde concentrations determined with the passive aldehydes and ketones sampler (PAKS) method (36, 37); however, the acetaldehyde (and other carbonyls) concentrations determined by DNPH-coated solid sorbents appear to be consistently and substantially lower than the acetaldehyde concentrations determined with the PAKS method (36, 37). In addition, Lazarus (25) used a dynamic dilution system to evaluate the 12- and 48-hour CEs of DNPHcoated solid sorbents, and observed a trend similar to what has been observed in the current study. In his work, Lazarus (25) observed a substantially lower (by about 3 fold) CE for acetaldehyde (and for acetone) than for formaldehyde. These independent results support our observation that DNPHcoated solid sorbents have low CEs for long-term sampling of carbonyls other than formaldehyde. Implications and Recommendations. Results from this study have confirmed the findings of other studies that support the short-term sampling of formaldehyde and acetaldehyde, and the long-term sampling of formaldehyde with DNPH-coated solid sorbent sampling methods. However, results from our long-term sampling experiments indicate that DNPH-coated solid sorbent sampling methods have acetaldehyde CEs that are substantially less than 100% under our experimental conditions. Our observations, along with those of Grosjean and Grosjean (21) and Kleindienst et al. (23), clearly demonstrate that there can be discrepancies with DNPH-coated solid sorbent sampling methods that are not explained by cartridge breakthrough, although the reaction mechanisms need to be further investigated to explain and “remediate” the observed discrepancies. Assuming that DNPH-coated solid sorbent methods have ∼100% acetaldehyde CEs for long-term sampling, simply based on breakthrough results or short-term sampling evaluations, will result in a substantial under-estimation of acetaldehyde concentrations. This is likely to be the case for other carbonyls as well; and evaluations of long-term sampling CEs, using DNPH-coated solid sorbents, needs to be extended to other commonly measured carbonyls in future studies.

Acknowledgments We thank Shamayne Cumberbatch for her assistance in this project. We also thank the Harvard School of Public Health (HSPH) for their assistance in the inter-laboratory quality. Jason Herrington was supported by a fellowship from the U.S. Department. of Education, Graduate Assistance in Areas of National Need, award no. P200A010808. The work of Drs. Zhang, Fan, and Lioy is supported, in part, by NIEHS Center grant no. 05022-10.

Literature Cited (1) U.S. EPA. Evaluation Report: Progress Made in Monitoring Ambient Air Toxics, but Further Improvements Can Increase Effectiveness; U.S. Environmental Protection Agency: Washington, DC, 2005. (2) U.S. EPA. Determination of Formaldehyde in Ambient Air Using Absorbent Cartridge Followed by High Performance Liquid Chromatography (HPLC) [active sampling methodology]: Compendium Method TO-11A in Compendium of Methods for the Determination of Toxic Organic Compounds in Ambient Air; U.S. Environmental Protection Agency: Washington, DC, 1999. (3) U.S. EPA. Final Draft: National Monitoring Strategy Air Toxics Component; U.S. Environmental Protection Agency: Washtington, DC, 2004.

(4) Andersson, G.; Andersson, K.; Nilsson, K.; Levin, J. O. Chemosorption of formaldehyde on Amberlite XAD-2 coated with 2,4dinitrophenylhydrazine. Chemosphere 1979, 8, 823-827. (5) Beasley, R. K.; Hoffmann, C. E.; Rueppel, M. L.; Worley, J. W. Sampling of formaldehyde in air with coated solid sorbent and determination by high performance liquid chromatography. Anal. Chem. 1980, 52, 1110-1114. (6) Andersson, K.; Hallgren, C.; Levin, J. O.; Nilsson, C. A. Solid chemosorbent for sampling sub-ppm levels of acrolein and glutaraldehyde in air. Chemosphere 1981, 10, 275-280. (7) Andersson, G.; Andersson, K.; Nilsson, K.; Levin, J. O. Chemosorption sampling and analysis of formaldehyde in air. Scand. J. Work. Environ. Health 1981, 7, 282-289. (8) Grosjean, D.; Fung, K. Collection efficiencies of cartridges and microimpingers for sampling of aldehydes in air as 2,4dinitrophenylhydrazones. Anal. Chem. 1982, 54, 1221-1224. (9) Kuwata, K.; Uebori, M.; Yamasaki, H.; Kuge, Y. Determination of aliphatic aldehydes in air by liquid chromatography. Anal. Chem. 1983, 55, 2013-2016. (10) Lipari, F.; Swarin, S. J. 2,4-dinitrophenylhydrazine-coated florisil sampling cartridges for the determination of formaldehyde in air. Environ. Sci. Technol. 1983, 19, 70-74. (11) Levin, J. O.; Andersson, K.; Lindahl, R.; Nilsson, C. A. Determination of sub-part-per-million levels of formaldehyde in air using active or passive sampling on 2,4-dinitrophenylhydrzinecoated glass fiber filters and high-performance liquid chromatography. Anal. Chem. 1985, 57, 1032-1035. (12) Tejada, S. B. Evaluation of silica gel cartridges coated in situ with acidified 2,4-dinitrophenylhydrazine for sampling aldehydes and ketones in air. Int. J. Environ. Anal. Chem. 1986, 26, 167-185. (13) Arnst, R. R.; Tejada, S. B. 2,4-dinitrophenylhydrazine-coated silica gel cartridge method for determination of formaldehyde in air: identification of an ozone interference. Environ. Sci. Technol. 1989, 23, 1428-1430. (14) Fung, K.; Wright, B. Measurement of formaldehyde and acetaldehyde using 2,4-dinitrophenylhydrazine-impregnated cartridges during the carbonaceous species methods comparison study. Aerosol Sci. Technol. 1990, 12, 44-48. (15) Druzik, C. M.; Grosjean, D.; Van Neste, A.; Parmar, S. S. Sampling of atmospheric carbonyls with small DNPH-coated C18 cartridges and liquid chromatography analysis with diode array detection. Int. J. Environ. Anal. Chem. 1990, 38, 495-512. (16) Zhou, X.; Mopper, K. Measurement of sub-parts-per-billion levels of carbonyl compounds in marine air by a simple cartridge trapping procedure followed by liquid chromatography. Environ. Sci. Technol. 1990, 24, 1482-1485. (17) Grosjean, D. Ambient levels of formaldehyde, acetaldehyde, and formic acid in outhern California: results of a one-year base-line study. Environ. Sci. Technol. 1991, 25, 710-715. (18) Slemr, J. Determination of volatile carbonyl compounds in clean air. Fresenius’ J. Anal. Chem. 1991, 340, 672-677. (19) Sirju, A. P.; Shepson, P. B. Laboratory and field investigation of the DNPH cartridge technique for the measurement of atmospheric carbonyl compounds. Environ. Sci. Technol. 1995, 29, 384-392. (20) Grosjean, E.; Grosjean, D. Performance of DNPH-coated C18 cartridges for sampling c1-c9. carbonyls in air. Int. J. Environ. Anal. Chem. 1995, 61, 343-360. (21) Grosjean, E.; Grosjean, D. Carbonyl collection efficiency of the DNPH-coated C18 cartridge in dry air and in humid air. Environ. Sci. Technol. 1996, 30, 859-863. (22) Goelen, E.; Lambrechts, M.; Geyskens, F. Sampling intercomparisons for aldehydes in simulated workplace air. Analyst 1997, 122, 411-419.

(23) Kleindienst, T. E.; Corse, E. W.; Blanchard, F. T. Evaluation of the performance of DNPH-coated silica gel and C18 cartridges in the measurement of formaldehyde in the presence and absence of ozone. Environ. Sci. Technol. 1998, 32, 124-130. (24) Binding, N.; Schilder, K.; Czeschinski, P. A.; Witting, U. Simultaneous determination of airborne acetaldehyde, acetone, 2-butanone, and cyclohexanone using sampling tubes with 2,4dinitrophenylhydrazine-coated solid sorbent. Toxicol. Lett. 1998, 96-97., 289-299. (25) Lazarus, E. Ph.D. Thesis, Rutgers, The State Univserity of New Jersey, NJ, 1999 (26) Possanzini, M.; Di, Palo, V.; Brancaleoni, E.; Frattoni, M.; Ciccioli, P. A train of carbon and DNPH-coated cartridges for the determination of carbonyls from C1 to C12 in air and emission samples. Atmos. Environ. 2000, 34, 5311-5318. (27) Sandner, F.; Dott, W.; Hollender, J. Senstive indoor air monitoring of formaldehyde and other carbonyl compounds using the 2,4dinitrophenylhydrazine method. Int. J. Hyg. Environ. Health 2001, 203, 275-279. (28) Ho, S. S. H.; Yu, J. Z. Determination of airborne carbonyls: comparison of a thermal desorption/GC method with the standard DNPH/HPLC method. Environ. Sci. Technol. 2004, 38, 862-870. (29) Fung, K.; Grosjean, D. Determination of nanogram amounts of carbonyls as 2,4-dinitrophenylhydrazones by high-performance liquid chromatography. Anal. Chem. 1981, 53, 168-171. (30) Vairavamurthy, A.; Roberts, J. M.; Newman, L. Methods for determination of low molecular weight carbonyl compounds in the atmosphere: a review. Atmos. Environ. 1992, 26A, 19651993. (31) IARC. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans; International Agency for Research on Cancer: Lyon, France, 2004. (32) Cogliano, V. J.; Grossem, Y.; Baan, R. A.; Straif, K.; Secretan, M. B.; El, Ghissassi, F. Meeting report: summary of IARC monographs on formaldehyde, 2-butoxyethanol, and 1-tert-butoxy2-propanol. Environ. Health Perspect. 2005, 113, 205-208. (33) U.S. EPA. Integrated Risk Information System (IRIS) on Acetaldehyde; U.S. Environmental Protection Agency: Washington, DC, 2006. (34) Zhang, J.; He, Q.; Lioy, P. J. Characteristics of aldehydes: concentrations, sources, and exposures for indoor and outdoor residential microenvironments. Environ. Sci. Technol. 1994, 28, 146-152. (35) Zhang, J.; Smith, K. R. Emissions of carbonyl compounds from various cookstoves in china. Environ. Sci. Technol. 1999, 33, 2311-2320. (36) Zhang, J.; Zhang, L.; Fan, Z.; Ilacqua, V. Development of the personal aldehydes and ketones sampler based upon DNSH derivatization on solid sorbent. Environ. Sci. Technol. 2000, 34, 2601-2607. (37) Herrington, J.; Zhang, L.; Whitaker, D.; Sheldon, L.; Zhang, J. J. Optimizing a dansylhydrazine (DNSH) based method for measuring airborne acrolein and other unsaturated carbonyls J. Environ. Monit. 2005, 7, 969-976. (38) Relationships of Indoor, Outdoor, and Personal Air (RIOPA), Part I, Collection Methods and Descriptive Analyses; Health Effects Institute: Boston, 2005.

Received for review May 24, 2006. Revised manuscript received September 26, 2006. Accepted October 20, 2006. ES061247K

VOL. 41, NO. 2, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

585